SUMMARY/ABSTRACT Whole genome and exome sequence data is currently being generated for population- and family-based studies to elucidate the involvement of rare variants in the etiology of complex traits including on Alzheimer?s Disease (AD). Although many rare variant population-based association methods have been developed there are extremely few methods to study families. Analyzing families to detect complex trait associations can be advantageousbecausesusceptibilityvariantsthatsegregateinfamiliescanhavelargereffectsizesthanthose found in sporadic cases, thereby increasing the power for detection, while avoiding spurious findings due to populationsubstructureandadmixture,thatcanplaguerarevariantpopulation-basedstudies.Forrarevariant complextraitanalysis,wewilldevelopfamily-basedassociationandlinkagemethods.Rarevariantmixedmodel association methods will also be developed for analysis of related and unrelated individuals. All developed methodswillbeusedtostudylate-onsetAD;?analyzingwholegenomeandexomesequencedatageneratedon families, cases and controls to discover novel genes and elucidate mechanisms underlying AD. AD status, as wellasquantitativetraitsageofonset,memoryandmemorydeclinewillbeanalyzed.Thedevelopedmethods willbeimplementedinourSEQSparksoftwaretoallowforrapidanalysisthroughparallelprocessing.Completion ofthisstudywilldevelopmethodsandsoftwaretoelucidatecomplextraitetiology.Applicationofthesemethods, analyzingexistingADsequencedatafromtheAlzheimer?sDiseaseSequencingProject(ADSP)andtheNational InstituteofAgingLate-onsetAlzheimer?sDisease(NIALOAD)study,willelucidateabetterunderstandingoflate- onsetADetiologyandriskfactors.IdentifyingsusceptibilityvariantsforADisthefirststepinriskpredictionand developmentoftreatmentswithhighefficacy.Thisstudyhashighpublichealthsignificance,sincelate-onsetAD causesconsiderablemorbiditywithintheelderlyandADprevalenceisincreasingduetoanagingpopulation.